I must disagree. As an organization grows, the authority structure must be made simple enough to understand or it won't work. N-ary trees are really simple, and with good people they work.
I once watched an organization struggle for months to come up with an explanation of the structure they were trying to transition to. Model after model were discarded. I eventually realized that the problem was that every effective model looked like a tree, and they *knew* they didn't want a tree, so out went *that* model.
Commercial software's motivators used to be different. You could just buy version N and stick with it forever, with no support beyond a ten-day warranty period. Or you could subscribe to "software maintenance", which gave you rights to expect a certain level of responsiveness whenever you had a problem *and* also helped to fund further development. With a steady revenue stream from maintenance fees, they could afford to work more nearly on a "when it's ready" basis.
Nowadays there's no revenue from quick fixes; the only revenue is from the initial sale. So theres a relentless pressure to ship something, ANYTHING, to keep the sales going. If you don't crank out the versions often enough, you could run out of money because the market's full of what you have today and you have no other source of income.
Some vendors offer per-incident paid support, but my experience is that that never gets used, because of the hassle involved in getting approval for unplanned expenditures. Figuring out a workaround in-house may cost more, but the cost isn't so easy to see and there's no delay while someone ponders whether to spend it.
The scheduling problem that many FOSS projects have is that no scheduling is ever done. Nobody is making a list, saying, "okay, these features will be accepted into 2.2; those, and anything else from now on, go into 2.4." So there are umpty-ump delays before 2.2 (whatever the heck that means) ships.
Then there are the projects that never seem to release anything; they go on for years at release 0.3, and whenever someone complains, the answer is, "oh, that is already in HEAD". Random snapshots of experimental code do not a release make.
"Many eyes" is good, but it's the scheduling disaster that works against producing solid designs and solid code in the first place. Linux vN+1 ships when Linus thinks it's ready, not to meet some marketing manager's fantasy deadline. Shipping software when it's ready, not when someone who hadn't a thing to do with making it wishes it would be ready, cannot be overvalued as a component of software quality.
Remember 'way back during the Cold War, when we in the U.S. thought that the Soviet Union was rather silly to arrest people just for taking pictures of things anyone could see, and making the most innocuous documents secret?
Whether the two parties are doing equal work is central to the discussion of whether NASA ought to be embarassed that a short suborbital flight done decades after they did it was done for substantially less money.
Whether NASA has failed to make progress after 1969 depends on which direction you think "forward" is. Is "put people into space cheaply" in their charter? What do you think they were supposed to be doing? Does the Congress agree with you?
That's Not NASA's Job. NASA is *supposed* to be a stepping stone. There's no point in them doing Alan Shepard's flight again in 2004 because private concerns are now able to do (almost) that much. NASA got us far enough for that to happen and is supposed to hand stuff off to industry when it can be routinized. I'm sure that if someone saw profit in building and maintaining a private copy of STS, it would happen and NASA would be out of the shuttle business.
Today's event is not a sign of failure for public space programs, but a sign of their success.
The argument that I'm making is that it is not reasonable to compare the arrival times of runners who don't start at the same distance from the finish line. Of course one can do a smaller job more cheaply. The runner who had the longer course need feel no embarassment at having taken longer to reach the goal. Likewise an organization which had to invent manned space flight from zilch will have had to spend more than one which could just go to the library and look up most of the science. Starting with the engineering saves loads of money compared to starting with the science so that you can then do the engineering.
Could NASA build a suborbital man-carrier for $25 million? Probably not, because they *are* a big bureaucracy. Could they do it for $50 million? Maybe. Is there any point to them doing it at all? No, because they passed that point decades ago and now private concerns are able to take the job on, leaving NASA to do its job: take on the missions that no private concern wants to pay for, but which will (we may hope) pave the way for later private uses of space.
Dunno, where is NASA's current mandate to toss something just over the 100km limit for immediate return? Someday the commercial space industry will look back on today and reflect that it had to be done, it made the industry possible, but there sure isn't any point in doing *that* again.
Indeed, you should see all of the information (*not* just PR) that NASA publishes, for anyone to read, compliments of the U.S. taxpayer. You're welcome.
*sigh* But the people who built the X-15 didn't have an X-15 and oodles of flight data from it to look at. It's a whole lot harder (and more expensive) the first time.
Half a century from now, when public-funded Mars missions have paved the way, I'm sure that private Mars trips will cost a lot less than $10Xbillion. But not today.
An important point zoomed by here: anyone who is worried about practicality should take a look at any convention center or big-name hotel and realize that practicality is the only thing they left out. Nobody is going to build a space hotel to be practical, for decades at least; they'll build them because it's cool enough to rake in big bucks from the tourists. People who actually want to *work* in space will happily live in a tin can just for the opportunity, and many have. They should have decent accommodation, and we can do a lot in this area as our capabilities mature, but nobody is going to be putting up construction workers or scientists in the Hilton Interplanetary.
But I must object to "embarrassingly smaller budget than NASA's." NASA had to do their first manned suborbital flight with 1950s hardware borrowed from the artillery boys, and without 40 years of prior experience to draw on.
The X Prize contestants are, in Newton's words, standing on the shoulders of giants. They're doing great things, and I applaud them, but there's no need to tear down other pioneers to build these guys up. The present work is quite impressive enough as it is.
Ah, I understand. "Brother sells its customers a new unsatisfactory printer every year -- we need to decrease quality until we can do that too! Our legendary reliability is killing sales."
If you have a good product, and you actually go out and present it to people who could use it, you can sell enough to keep it going and make some money. Changing the CPU under what people are doing is a really hard sell, I'll admit, but the various companies that had the AxP at one time or another never even tried, so far as I can tell.
Having a great product is not enough. The buyers have to know it's great, and sometimes that means you have to go out and tell them about it.
Slashdot Mirror
The difference is that there doesn't seem to be a way to design oil that changes itself.
I must disagree. As an organization grows, the authority structure must be made simple enough to understand or it won't work. N-ary trees are really simple, and with good people they work.
I once watched an organization struggle for months to come up with an explanation of the structure they were trying to transition to. Model after model were discarded. I eventually realized that the problem was that every effective model looked like a tree, and they *knew* they didn't want a tree, so out went *that* model.
Commercial software's motivators used to be different. You could just buy version N and stick with it forever, with no support beyond a ten-day warranty period. Or you could subscribe to "software maintenance", which gave you rights to expect a certain level of responsiveness whenever you had a problem *and* also helped to fund further development. With a steady revenue stream from maintenance fees, they could afford to work more nearly on a "when it's ready" basis.
Nowadays there's no revenue from quick fixes; the only revenue is from the initial sale. So theres a relentless pressure to ship something, ANYTHING, to keep the sales going. If you don't crank out the versions often enough, you could run out of money because the market's full of what you have today and you have no other source of income.
Some vendors offer per-incident paid support, but my experience is that that never gets used, because of the hassle involved in getting approval for unplanned expenditures. Figuring out a workaround in-house may cost more, but the cost isn't so easy to see and there's no delay while someone ponders whether to spend it.
The scheduling problem that many FOSS projects have is that no scheduling is ever done. Nobody is making a list, saying, "okay, these features will be accepted into 2.2; those, and anything else from now on, go into 2.4." So there are umpty-ump delays before 2.2 (whatever the heck that means) ships.
Then there are the projects that never seem to release anything; they go on for years at release 0.3, and whenever someone complains, the answer is, "oh, that is already in HEAD". Random snapshots of experimental code do not a release make.
"Many eyes" is good, but it's the scheduling disaster that works against producing solid designs and solid code in the first place. Linux vN+1 ships when Linus thinks it's ready, not to meet some marketing manager's fantasy deadline. Shipping software when it's ready, not when someone who hadn't a thing to do with making it wishes it would be ready, cannot be overvalued as a component of software quality.
Remember 'way back during the Cold War, when we in the U.S. thought that the Soviet Union was rather silly to arrest people just for taking pictures of things anyone could see, and making the most innocuous documents secret?
Meet Mr. Mirror.
...was a *radio*telescope. :-/
Whether the two parties are doing equal work is central to the discussion of whether NASA ought to be embarassed that a short suborbital flight done decades after they did it was done for substantially less money.
Whether NASA has failed to make progress after 1969 depends on which direction you think "forward" is. Is "put people into space cheaply" in their charter? What do you think they were supposed to be doing? Does the Congress agree with you?
That's Not NASA's Job. NASA is *supposed* to be a stepping stone. There's no point in them doing Alan Shepard's flight again in 2004 because private concerns are now able to do (almost) that much. NASA got us far enough for that to happen and is supposed to hand stuff off to industry when it can be routinized. I'm sure that if someone saw profit in building and maintaining a private copy of STS, it would happen and NASA would be out of the shuttle business.
Today's event is not a sign of failure for public space programs, but a sign of their success.
The argument that I'm making is that it is not reasonable to compare the arrival times of runners who don't start at the same distance from the finish line. Of course one can do a smaller job more cheaply. The runner who had the longer course need feel no embarassment at having taken longer to reach the goal. Likewise an organization which had to invent manned space flight from zilch will have had to spend more than one which could just go to the library and look up most of the science. Starting with the engineering saves loads of money compared to starting with the science so that you can then do the engineering.
Could NASA build a suborbital man-carrier for $25 million? Probably not, because they *are* a big bureaucracy. Could they do it for $50 million? Maybe. Is there any point to them doing it at all? No, because they passed that point decades ago and now private concerns are able to take the job on, leaving NASA to do its job: take on the missions that no private concern wants to pay for, but which will (we may hope) pave the way for later private uses of space.
Dunno, where is NASA's current mandate to toss something just over the 100km limit for immediate return? Someday the commercial space industry will look back on today and reflect that it had to be done, it made the industry possible, but there sure isn't any point in doing *that* again.
Indeed, you should see all of the information (*not* just PR) that NASA publishes, for anyone to read, compliments of the U.S. taxpayer. You're welcome.
Yeah, you just take the Pan Am shuttle up to orbit, then transfer to an Aries 1B for the remainder of the trip.
When you arrive, please do not disturb the monoliths.
*sigh* But the people who built the X-15 didn't have an X-15 and oodles of flight data from it to look at. It's a whole lot harder (and more expensive) the first time.
Half a century from now, when public-funded Mars missions have paved the way, I'm sure that private Mars trips will cost a lot less than $10Xbillion. But not today.
Bandwidth, shmandwidth. It's Windows Media only. And their silly Flash controls don't work here either.
An important point zoomed by here: anyone who is worried about practicality should take a look at any convention center or big-name hotel and realize that practicality is the only thing they left out. Nobody is going to build a space hotel to be practical, for decades at least; they'll build them because it's cool enough to rake in big bucks from the tourists. People who actually want to *work* in space will happily live in a tin can just for the opportunity, and many have. They should have decent accommodation, and we can do a lot in this area as our capabilities mature, but nobody is going to be putting up construction workers or scientists in the Hilton Interplanetary.
Okay, way to go Scaled team!
But I must object to "embarrassingly smaller budget than NASA's." NASA had to do their first manned suborbital flight with 1950s hardware borrowed from the artillery boys, and without 40 years of prior experience to draw on.
The X Prize contestants are, in Newton's words, standing on the shoulders of giants. They're doing great things, and I applaud them, but there's no need to tear down other pioneers to build these guys up. The present work is quite impressive enough as it is.
Ah, I understand. "Brother sells its customers a new unsatisfactory printer every year -- we need to decrease quality until we can do that too! Our legendary reliability is killing sales."
Elron is visiting relatives on Tol Eressea this month. Maybe you mean L. Ron?
So, did they hypnotize me into believing that the WTC is not there now, or that it was there before?
Wait, I know! They did BOTH!!! OMG we are *owned*!
We are not shy about kicking manufacturers for unwelcome decisions, so here's praise for one that has rethought and made a better decision.
If you have a good product, and you actually go out and present it to people who could use it, you can sell enough to keep it going and make some money. Changing the CPU under what people are doing is a really hard sell, I'll admit, but the various companies that had the AxP at one time or another never even tried, so far as I can tell.
Having a great product is not enough. The buyers have to know it's great, and sometimes that means you have to go out and tell them about it.
Sounds like the disease that killed DEC, then spread to Compaq, is now chewing happily on HP. Sad.
What technology *is* UDC, exactly? I followed a couple of links but couldn't find anything to explain why I should want it.